亚稳多晶型药物的应用前景及人工智能在该领域的研究进展  

Application prospects of metastable polymorph drugs and the research progress of artificial intelligence

作　　者：万斌[1] 于翔 丁泽杰 Wan Bin;Yu Xiang;Ding Zejie(Department of Pharmacy,Huzhou Central Hospital,Huzhou,Zhejiang 313000,China;PhaseⅠClinical Trial Center,Huzhou Third Municipal Hospital,Huzhou,Zhejiang 313000,China)

机构地区：[1]湖州市中心医院药学部,浙江湖州313000 [2]湖州市第三人民医院药物Ⅰ期临床研究室,浙江湖州313000

出　　处：《中国药物与临床》2025年第3期197-204,F0003,共9页Chinese Remedies & Clinics

基　　金：浙江省湖州市公益性应用研究项目(2023GY11)。

摘　　要：亚稳晶型广泛存在于各类固态药物中,研究药物亚稳晶型对于药学领域的发展至关重要。本文总结了亚稳晶型相较于稳定晶型在压片性能、引湿性、体外溶出、体内药理活性等方面的优越性。详细介绍了多种可能应用于亚稳晶型成药的固体药物递送系统,以及基于“固溶体热力学开关”逆转亚稳晶型稳定性的方法学,这些方法为亚稳晶型的成药拓宽了新的思路。此外,本文系统地阐述了人工智能在亚稳晶型领域的应用前景,通过随机森林模型、机器学习技术、数字化统计等方法可能可以预测、解析晶体结构或制备新的亚稳晶型,在未来也能预测新晶型的理化性质。Metastable polymorphs are widely present in various solid-state drugs,and it is crucial for the advancement of pharmaceutical science.This article summarizes the advantages of metastable polymorphs over stable polymorphs in terms of tabletability,hygroscopicity,in vitro dissolution,and in vivo pharmacological activity.It also provides a detailed introduction to various solid drug delivery systems that may be applied to the development of metastable polymorphs which broaden new avenues for drug development,as well as the methodology based on the“thermodynamic switch of solid solutions”to reverse the stability of metastable polymorphs.Finally,the article systematically discusses the prospects of artificial intelligence in the field of metastable polymorphs.Through methods such as random forest models,machine learning techniques and digital statistics,it may be possible to predict and analyze the crystal structures and prepare new metastable polymorphs,potentially allowing for the prediction of new polymorphs’physicochemical properties in the future.

关 键 词：亚稳晶型 药物传递系统 人工智能 

分 类 号：R943[医药卫生—药剂学]